Health Care Worker (HCW) performance in low-resource settings is frequently below the standards of care recommended in Simplified Clinical Practice Guidelines (SCPGs), leading to millions of unnecessary deaths each year. Performance, defined as adherence to a standard, depends upon appropriate training and continuous learning. Training interventions to improve HCW performance are based on SCPGs that standardize the delivery of healthcare. Once training is complete, HCWs have few opportunities to maintain and improve their knowledge through continuous learning by, for example, receiving feedback that reinforces new guideline-based knowledge. A barrier to the provision of feedback in low-resource settings is a shortage in human resources. However, the increased use of Electronic Medical Record (EMR) systems in low-income countries is creating new opportunities to generate automated feedback that can support learning, thereby improving HCW performance. In a preliminary study, we determined the feasibility of using EMR data to generate automated feedback about Malawi's national guidelines for the treatment of AIDS. Using a conceptual model of employee performance feedback, we propose to (1) identify HCW perceptions and contextual factors affecting implementation of automated audit and feedback in Malawi, and (2) develop a prototype automated audit and feedback system to generate performance feedback about AIDS treatment guidelines. This work will extend the existing healthcare infrastructure in Malawi to provide routine, automated feedback by reusing EMR data at low cost to the health care organization. PUBLIC HEALTH RELEVANCE: Healthcare worker performance in low-resource settings is frequently below the standards of care put forth in clinical practice guidelines, leading to millios of unnecessary deaths each year. Performance depends upon appropriate training and continuous learning. Healthcare workers have limited opportunities to improve their knowledge by, for example, receiving feedback that reinforces new guideline-based knowledge. However, the arrival of electronic health information systems is creating new opportunities to generate automated feedback that can support learning, thereby improving health care worker's performance. In this study we propose to (1) determine the environmental and cultural factors affecting implementation of such learning systems, and (2) develop a prototype automated audit and feedback system to generate guideline adherence feedback in the domains of both the treatment of AIDS and diabetes care.